leaderboard / sharegpt /README.md
Jae-Won Chung
Add benchmark.py, example data, README.md
a679cf2
|
raw
history blame
957 Bytes

How we used ShareGPT to create our benchmark dataset

Download ShareGPT

https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/HTML_cleaned_raw_dataset/sg_90k_part1_html_cleaned.json

https://huggingface.co/datasets/anon8231489123/ShareGPT_Vicuna_unfiltered/resolve/main/HTML_cleaned_raw_dataset/sg_90k_part2_html_cleaned.json

Install Fastchat

pip install fastchat

Clean data:

pip install polyglot pyicu pycld2
python -m fastchat.data.optional_clean --in sg_90k_part1_html_cleaned.json --out sg_90k_part1_html_cleaned_lang.json --keep-lang en

Extract first prompt

python extract_first.py --in-file sg_90k_part1_html_cleaned_lang.json --out-file sg_90k_part1_html_cleaned_lang_first.json

Sample data

python -m fastchat.data.sample --in sg_90k_part1_html_cleaned_lang_first.json --out sg_90k_part1_html_cleaned_lang_first_sampled.json --end 10000 --max-length 10000